Table 3.

Maximum Likelihood Parameter Estimates of Fixed Effects from Mixed Regression Models that Quantify (1) the Influence of Perceived Image Quality Ratings on Relative Acuity and (2) the Influence of Perceived Contrast, Blur, and Ghosting (Measured as Position Offset) on Overall Perceived Image Quality After Controlling for the Mean Effects of Metric Type

Characteristic	Estimate	Standard Error	95% Confidence Intervals	P Value
(1) Relative acuity*
Overall perceived image quality	0.001	0.00014	0.0006 to 0.0012	<0.0001
Perceived blur	–0.012	0.001	–0.014 to –0.005	<0.0001
Perceived ghosting	–0.003	0.001	–0.001 to –0.009	0.0053
Perceived contrast	0.002	0.001	–0.000 to 0.004	0.1285
(2) Overall perceived image quality**
Perceived blur	–4.37	0.13	–4.64 to –4.11	<0.0001
Perceived ghosting	–2.17	0.18	–2.40 to –1.94	<0.0001
Perceived contrast	–2.56	0.14	–2.84 to –2.28	<0.0001

^*Difference between clear chart acuity and experimental derived chart acuity.

^**Sensitivity analysis for model 2: bootstrap means and confidence intervals: blur (–4.72, –3.78), ghosting (–2.50, –1.77), and contrast (–2.93, –1.89).

Note: Additionally, deviance χ² statistics for the overall effects of blur, ghosting, and contrast were assessed in unconstrained models, with predictors entered categorically, comparing reduced and full models (i.e., Δ_−2LL=[−2LL(Reduced)− −2LL(Full)]∼χ² (9)). Estimated overall effects for model 1 were as follows: blur (Δ_−2LL=92.9), ghosting (Δ_−2LL=25.4), and contrast (Δ_−2LL=10.1). Estimated overall effects for model 2 were as follows: blur (Δ_−2LL=880.4), ghosting (Δ_−2LL318.2), and contrast (Δ_−2LL=357.7). All except the effect of contrast in model 1 surpass the .05 critical value, χ²(9) = 16.92 (P < 0.05) (i.e., P<0.05 when Δ_−2LL>16.92).